Novel displacement sensing : towards robotic tunnelling.

This thesis is presented in two parts. It was noted that in the second half of the 1990's, a period during which there were long and dry summers in the UK, the water utility companies were openly criticized for their inefficiency and poor distribution-network performance as supplies ran low. It was noted that some of these companies were suffering up to 50% loss of water through leaks in the distribution network. Increasing pressure mounted for the owners of buried infrastructure to more effectively identify and service both small and large leaks alike. Conventional methods are shown to be not effective or efficient in the long-term, and that it would be most useful to adopt a more appropriate strategy in the access to and repair of damaged water mains. A conceptual solution is developed and proposed about a 'minimal-intrusion' system which would allow efficient access to buried infrastructure for the purposes of repair, whilst reducing effects on the local enviromnent when compared to conventional excavation. The concept is based around biological mimicry as the highly adaptable access vehicle copies both the form of and the locomotion strategy of a common earthworm. The associated technological challenges are identified and it is realized that to achieve a technical content within the project it is necessary to focus on one particular area of interest. The area selected is locomotion of a worm-like vehicle. Suitable actuators are discussed in some detail, as is a method for deriving feedback on the surface form of the vehicle, which may vary greatly according to conditions. A justification is made to pursue the development of a surface mounted or embedded sensor network which is able to profile the form of a complex surface. The sensor technology is selected after some consideration, and is based on inductive principles. A completely novel sensor in terms of form and electronic/signal-processing design is developed from first principles. A prototype is constructed from suitable materials and subjected to a considered testing strategy. The testing is based around a custom built rig, which is intended to allow extended duration procedures and automated data acquisition. The prototype is subjected to several iterations of refinement before a main test series is conducted. The performance of the prototype is compared with a conventional sensing device where applicable. Finally, some thought is given to the application of the new sensor in the context of this project, and in other fields.